This research project uses novel and innovative strategies to develop high-fidelity optic nerve head (ONH) structure endophenotypes computationally. These computational phenotypes allow quantification of ONH structure, and will be measured for each image, spanning multiple imaging modalities from several well-characterized cohorts (OHTS, AREDS, IOWA, and Rotterdam). We will use a genetic imputation strategy to maximize our power to identify bona fide associations, and ultimately the causative variations, for our computational ONH structure endophenotypes. A systems biology approach will be employed to identify biological and/or functional relationships among the associated loci and genes. The ultimate goal of this research proposal is identify biomarkers and/or genetic risk factors that accurately predict: (1) primary optic nerve head (ONH) structure (i.e. before age- or disease-related changes), (2) changes in ONH structure, and (3) the development of irreversible glaucomatous optic nerve damage before it occurs. These outcomes will improve the specificity and sensitivity of initial diagnosis of glaucoma, allowing clinicians to determine the proportion o ONH structure change that is damage from this disease, as opposed to normal variations in primary ONH structure. This in turn will allow the application of currently available and effective therapies to be instituted before vision is lost.